A Laboratory Method for Measuring the Heat Distribution Of

A LABORATORY METHOD FOR MEASURING THE HEAT DISTRIBUTION OF

LUMINAIRES AND ITS APPLICATION FOR BUILDING

HEATING AND COOLING ENERGY SAVING

Hankun Li

Submitted to the graduate degree program in Architectural Engineering and the Graduate Faculty of the University of Kansas in partial fulfillment of the requirements for the degree of Master of Science.

______

Chairperson Dr. Hongyi Cai

______

Dr. Brian A. Rock

______

Dr. Xianglin Li

Date Defended:

The Thesis Committee for Hankun Li

certifies that this is the approved version of the following thesis:

A LABORATORY METHOD FOR MEASURING THE HEAT DISTRIBUTION OF

LUMINAIRES AND ITS APPLICATION FOR BUILDING

HEATING AND COOLING ENERGY SAVING

______

Chairperson Dr. Hongyi Cai

Date approved: <<

ABSTRACT

This study was aimed to explore the energy saving potential of a new system architecture of solid-state lighting fixtures that was designed to help utilize the heat generated by LEDs for spacing heating in heating season and re-heating in cooling season. The new system architecture, which deploys an innovation of integrative light and heat arrangement in low profile, helps harvest the LED heat and direct most of the heat to the room space while minimizing heat leakage to the ceiling cavity. A well-designed laboratory experiment was carried out in a newly developed

Calorimeter chamber that was used to find out heat distribution of luminaires in the conditioned room cavity and ceiling plenum, followed by an estimation of potential energy savings via computer simulation in Energy Plus. A typical primary elementary school classroom was used in this computer simulation equipped with LED fixtures with different heat distribution patterns. It was found that in heating season, the building space heating energy consumption could be reduced as the ‘conditioned space/ceiling plenum split’ increased. While in cooling season, the LED heat gain in the conditioned room could be utilized to warm up the chilled supply air to supplement the function of reheating system. The new system architecture of LED fixtures with integrative lighting and heating arrangement could save 4.4%-4.7% of annual building heating and cooling energy uses by reducing reheating energy consumption in cooling season and space heating energy consumption in heating season.

iii

ACKNOWLEDGMENTS

First of all, many thanks to Dr. Hongyi Cai, the chairperson of my committee, for his guidance of my master program through past three semesters at the University of Kansas. I have learned how to be a responsible researcher and benefited a lot from his instructions.

Also, I must extend my thanks to my committee members, Dr. Xianglin Li and Dr. Brian

Rock, who provided valuable academic supports to my thesis.

Thanks to Faran Mahlab, Reid M. Poling, Habib Arjmand Mazidi and Rui Cao for helping me with my pilot study experiments and cold-room experiments.

Another big thank to Simon Diederich, who built the calorimeter and his Master’s thesis provided technical supports to the experiment stage of my thesis.

Finally, I must thank my parents for providing me unconditional supports throughout my

Master’s degree study at the University of Kansas.

TABLE OF CONTENTS

ABSTRACT ...... iii

ACKNOWLEDGMENTS ...... iv

TABLE OF CONTENTS ...... v

LIST OF TABLES ...... vii

LIST OF FIGURES ...... ix

CHAPTER 1 INTRODUCTION ...... 1

1.1 Background ...... 1

1.1.1 Thermal Performance of LED luminaires ...... 1

1.1.2 Energy Plus Simulations ...... 7

1.1.3 Research Problem ...... 10

1.2 Research Goals and Objectives ...... 11

CHPATER 2 LITERATURE REVIEW ...... 16

2.1 Heat gains from space lighting ...... 16

2.2 Calorimeters ...... 20

2.3 Reheat systems ...... 21

CHAPTER 3 METHODOLOGY ...... 23

3.1 Theoretical Framework ...... 23

3.1.1 The calculation of Heating power and its distribution ...... 23

3.1.2 Research assumptions ...... 27

3.2 Pilot Study ...... 31

3.2.1 Darkroom experiments ...... 31

3.2.2 Data treatments and results ...... 35

3.2.3 Recommended improvements for formal experiment ...... 38

3.3 Experiments in the Cold Room ...... 39

3.3.1 Introduction ...... 39

3.3.2 Calibration ...... 43

3.3.3 Formal Experiments and results...... 48

3.4 Computer-Aided simulation ...... 53

3.4.1 Designs of Energy-plus simulation ...... 53

3.4.2 Introduction to the modeling with the integrated lighting arrangement design ... 56

3.4.3 Results of Energy Plus simulation ...... 58

CHAPTER 4 DATA ANALYSIS AND RESULTS ...... 61

4.1 Case 1: the surface mounted prototype LED luminaire vs. the ceiling recessed

prototype LED luminaire ...... 61

4.2 Case 2: the ceiling recessed prototype LED luminaire vs. a commercial ceiling recessed LED

luminaire ...... 62

4.3 Case 3: the surface mounted prototype LED luminaire vs. a commercial ceiling recessed LED

luminaire ...... 63

4.4 Case 4: the ceiling recessed prototype LED with energy saving design and assumptions ...... 64

CHAPTER 5 CONCLUSIONS AND DISCUSSIONS ...... 67

5.1 Conclusions ...... 67

5.2 Discussions ...... 70

REFERENCES ...... 72

APPENDIX

Calculation sheet of the heating power of the tested LED luminaires in the Calorimeter.....

LIST OF TABLES

Table 1 Fractions of selected overhead fluorescent luminaires ...... 19

Table 2 Configurations of 10 selected LED luminaires’ tests in pilot study ...... 32

Table 3 Calculation sheet of heating power of tested luminaires ...... 35

Table 4 Results from “dark room” experiments...... 37

Table 5 Results of the cartridge heater tests in the Calorimeter...... 46

Table 6 Guidelines of acceptable room temperature ...... 49

Table 7 “Cold room” luminaire tests configurations ...... 49

Table 8 Results of the ceiling recessed prototype LED luminaire testing ...... 51

Table 9 Results of the surface mounted prototype LED luminaire testing ...... 51

Table 10 Results of the ceiling recessed commercial LED luminaire testing ...... 52

Table 11 Results of the surface mounted commercial LED luminaire testing ...... 52

Table 12 Annual heating and cooling energy for the primary school classroom without

deployment of the new lighting arrangement ...... 59

Table 13 heating and cooling energy of the classroom installed with the prototype LED fixture

and with new lighting arrangement deployed ...... 60

Table 14 surface mounted prototype vs. ceiling recessed prototype ...... 62

Table 15 ceiling recessed prototype vs. ceiling recessed commercial LED luminaire ...... 63

Table 16 surface mounted prototype vs. surface mounted commercial LED luminaire ...... 64

Table 17 the surface mounted prototype LED luminaire with energy saving design (under

assumption C) vs. a surface mounted commercial LED luminaire ...... 65

vii

Table 18 the ceiling recessed prototype LED luminaire with energy saving design (under

assumption C) vs. a ceiling recessed commercial LED luminaire...... 66

viii

LIST OF FIGURES

Figure 1 Ideal electroluminescence process generating photons in the PN-Junction of solid-state

lighting fixture ...... 2

Figure 2 A new luminaire system architecture of “Heat Arrangement of LED Arrays in Low

Profile” to harvest both the light and the heat generated by the same LEDs via a mingled

path for lighting and heating uses ...... 4

Figure 3 Heated air diffuser with LED lights ...... 4

Figure 4 The Calorimeter in the darkroom of KU lighting research laboratory ...... 6

Figure 5 the prototype ceiling recessed LED fixture ...... 7

Figure 6 Primary school classroom modeling (Sketch Up make) ...... 9

Figure 7 Define and modify the modeling (Open Studio) ...... 9

Figure 8 a sample of Energy Plus simulation results ...... 10

Figure 9 The concept of integrative illumination and HVAC system, which integrate the

harvested LED heat in conditioned space with the air conditioning system ...... 14

Figure 10 Heat conversion for typical light sources ...... 17

Figure 11 Heat distribution of a fluorescent fixture ...... 18

Figure 12 A typical reheat system in commercial building ...... 22

Figure 13 Two types of sensor installed in the calorimeter with aluminum foil covered interior

surfaces ...... 28

Figure 14 Lights definitions setting page ...... 29

Figure 15 Differently mounted prototype LED tested in the calorimeter, (a) Ceiling recessed

prototype LED, (b) Surface mounted prototype LED ...... 32

Figure 16 The prototype LED luminaire with surface mounted design ...... 33

Figure 17 Data recording ...... 33

Figure 18 Temperature sensor at air intake of the room cavity ...... 34

Figure 19 The ceiling recessed prototype LED luminaire in the calorimeter ...... 34

Figure 20 Water vapor saturation pressure chart (www.engineeringtoolbox.com) ...... 36

Figure 21 The Cold Room laboratory, (a) the control panel, (b) a watch window on the door of

the Cold Room ...... 40

Figure 22 Setup of the calorimeter in the Cold Room ...... 41

Figure 23 Power cable and data wire ...... 42

Figure 24 Anti-condensation observation window ...... 42

Figure 25 Results of the first calibration: temperature increased from 4°C to 30°C (39°F to

86°F) with each step is 2°C...... 44

Figure 26 2nd Calibration of temperature sensors (21.0°C to 29.5°C) ...... 45

Figure 27 The pure heat source (cartridge heater) ...... 46

Figure 28 The Calibration of calorimeter heat loss at room temperature of 20°C ...... 47

Figure 29 The real time measurement of the controlled air-flow rate in ceiling cavity (red) and room cavity (purple) ...... 50

Figure 30 Floor plan and lights layout of the primary school classroom ...... 54

Figure 31 Define the heat distribution of luminaire ...... 55

Figure 32 Define the heating power of luminaire in ceiling plenum ...... 55

Figure 33 Define HVAC system of the modeling...... 56

Figure 34 Common working status of building space lighting and HVAC system ...... 57

Figure 35 Considered heat from luminaire (room portion) as a reheating source ...... 58

Figure 36 Summary of Energy Plus simulation ...... 59

CHAPTER 1

INTRODUCTION

1.1 Background

1.1.1 Thermal Performance of LED luminaires

Light-Emitted Diode (LED) lighting is a relatively mature technology today in

current lighting industry, which could provide the same amount of lumen output with much

less energy consumption (compared to conventional light sources). LED luminaires are

considered energy efficient and economical products because of their long life.

On the other hand, LED lighting has underestimated by-product of heat generation.

When LED generates light, 20%-30% of its electrical energy consumptions will be emitted

as visible light, and the rest 70%-80% of its consumed energy will be converted to heat

(U.S Department of Energy, 2008). Electroluminescence is the process in which LED chips

generate light and heat (Figure 1). In ideal electroluminescence process, when an electron

meets a hole, it falls into a lower energy level and releases energy by generating a photon.

Yet in most cases, electrons fallen into lower energy level will generate heat instead of

emitting photons. This heat will stay in the diode and cause temperature raising up quickly

at the back of the LED chips.

Figure 1 Ideal electroluminescence process generating photons in the PN-

Junction of solid-state lighting fixture (S-Kei, 2011)

To prevent LED chips from being burned out with overheat, a thermal heatsink is mounted on the back of LED to haul the heat stored on the back of LED chips away to cool the LED junction. The heatsink is made of materials with high thermal conductivity (e.g.

Aluminum, 118 Btu/hr °F ft) and often designed with a large surface area (e.g., with fins) to quickly dissipate heat to ambient environment through convection with ambient air.

Because most of the heat is generated on the back of LED that is quickly removed using heatsink, the temperature of the front surface of an LED light source is often only slightly higher than the ambient temperature. The highest operating temperature of LED chips is regulated as 120 °C (248 °F ) in current industrial standard (Hui, 2017).

Compared to conventional light sources, LEDs have two unique advantages that may change the lighting practice. First, conventional light sources such as incandescent bulbs or fluorescent tubes are with large size, often too big to be installed in small spaces

of built structures or to be integrated with building heating, ventilation and air conditioning

(HVAC) systems. In contrast, LED chips are tiny, can be installed in small spaces in low profile to enable innovative LED luminaires of next generation that integrate well with building systems. Second, the significant amount of heat generated by LED chips are accumulated and stored on the back of the LED, not radiated together with the light emitted in the forward direction, which is different from conventional light sources that generate light and heat in a mixed energy flux in the same direction. As a result, the heat generated by LED light source can be easily controlled and harvested without interfere its lighting performance (Cai, 2015).

Because of the two unique characteristics, a new luminaire system architecture of

“Heat Arrangement of LED Arrays in Low Profile” was developed by Cai (Cai, 2017,

“Integrated Light and Heat Arrangement of Low-Profile Light Emitting Diode Fixture”.

Non-provisional application No. 15/486,797, publication No. US-2017-0299167-A1) for integrative light and heat arrangement in low profile, as shown in Figure 2, from which future solid-state lighting fixtures could benefit. Via the new architecture of “Heat

Arrangement of LED Arrays in Low Profile”, the heat generated by LEDs can be effectively harvested and quickly dissipated into conditioned interior space for beneficial heating uses, without interruption of the LED light output. For example, LEDs in an array can be mounted on passive heat exchanger and integrated with lens in the front side and a layer of insulation materials on the back for making integrative LED fixtures in low profile

(Figure 2). Such slim LED fixtures can be ceiling recessed, surface mounted, or hung from the ceiling for integrative heating and lighting in conditioned interior spaces. Additionally,

LEDs can be attached to the metal frame of traditional supply air diffuser to make a light

and heat integrated air diffuser (Figure 3). It is thus expected that LED luminaires of next generation might be re-designed and re-engineered to integrate with the building HVAC system and maximize building energy savings in space heating, cooling, and lighting.

Figure 2 A new luminaire system architecture of “Heat Arrangement of LED Arrays in

Low Profile” to harvest both the light and the heat generated by the same LEDs via a

mingled path for lighting and heating uses (Cai, 2017).

Figure 3 Heated air diffuser with LED lights (Cai, 2015)

In fact, the impact of conventional lighting on building heating and cooling energy consumption has been well studied. Previous study (Fisher & Chantrasrisalai, 2007) on the impact of lighting on the building heating and cooling loads were focused on the integration of ceiling recessed lighting fixtures with air diffuser at the air return side. However, previous studies did not include newer LED lighting technologies, which are quite different from conventional lighting technologies in terms of size, light and heat generation, and real-time working performance. In addition to two unique characteristics of LED fixtures, there are other factors affecting the harvesting of light heat. For example, the ambient temperature affects the performance of fluorescent lights (e.g. Fluorescent T8, T12 lamps) which are designed to give maximum lights output at 25C (77°F). As a result, conventional fluorescent lamps would be dimmed in a cold ambient environment (e.g., 16

C at the supply air duct). However, LEDs do not have this problem, because LEDs work more efficiently in a cold environment due to LED heat being quickly removed to lower its junction temperature.

The present study was aimed to evaluate the thermal performance of custom- designed LED luminaires that deploy the new system architecture of “Heat Arrangement of LED Arrays in Low Profile” (Figure 2) in comparison to that of current LED fixtures available in the existing market. To that end, a portable calorimeter chamber (Figure 4) was designed by Dr. Hongyi Cai and constructed by Simon Diederich in the University of

Kansas Lighting Research Laboratory in a prior study (Diederich, 2017). The calorimeter was designed for accurate heat transfer measurement of new LED products in well- controlled environments for testing integrated solid-state lighting and heating technologies

of next generation. It has a size of approximately 2.6 ft wide, 6.5 ft long, and 6 ft high, divided into a ceiling space and a room space. The U value for the calorimeter walls is preset as 0.32 W/(m2 K) or less. In the present study, the calorimeter was used to measure thermal features (e.g., heating power and the heat distribution patterns) of selected testing

LED fixtures, including (i) prototypes of new surface-mounted or ceiling-recessed LED luminaires (Figure 5) that deploy the new system architecture of “Heat Arrangement of

LED Arrays in Low Profile” (test case), and (ii) two off-the-shelf commercial LED luminaires which were tested in the calorimeter under the same testing conditions for comparison (control cases).

Figure 4 The Calorimeter in the darkroom of KU lighting research laboratory

Figure 5 A prototype ceiling recessed LED fixture that deploy the new system architecture of

“Heat Arrangement of LED Arrays in Low Profile”

1.1.2 Energy Plus Simulation

Energy Plus is a building energy simulation software used by engineers, architects,

and researchers to simulate annual or seasonal energy consumptions for heating, cooling,

ventilation, and lighting in buildings (energyplus.net, 2018). Energy Plus has some notable

capabilities in building energy consumption simulation. The load categories and the HVAC

system in each conditioned space can be set up and modified, enabling basic needs of the

proposed computer-aided simulation in the present study to figure out how much heating

and cooling energy may be saved by the new LED fixtures with new system architecture

of “Heat Arrangement of LED Arrays in Low Profile”.

One key feature of the new LED luminaires (Figure 5) designed and developed in house is that they can harvest and dissipate more LED heat, which is otherwise trapped in the ceiling plenum, into the conditioned room space than current commercial LED luminaires do. During the heating season, the LED heat harvested in the room space is considered a supplemental heating power in Energy Plus simulation. In the cooling season, the air first chilled for moisture control (with air temperature below dew point) needs to be reheated by the re-heating coil of the HVAC system to a satisfying temperature (e.g., above

55°F) before the air is discharged into the room space. The LED heat harvested in the room space from those new LED luminaires is considered supplemental re-heating power, which may help re-heat the discharged air and thus lower the energy consumption of the reheating coil of the HVAC system (Cai, 2015).

The computer simulation in the present study used a model of a typical classroom

(28’×36’) in primary schools built by Sketch-up Make (Figure 6), with lighting configurations, climate data and other thermal parameters defined in Open studio (Figure

7). The model was then imported into Energy Plus for simulation of annual building energy consumption. Example simulation results are shown in Figure 8. The cooling and heating energy consumption data were extracted from the Energy Plus simulation report for further data analysis.

Figure 6 The modeling of the primary school classroom in Sketch Up make

Figure 7 Definition and modification of the modeling in Open Studio

Figure 8 An example of Energy Plus simulation results

1.1.3 Research Problems

There are two major research problems to solve in the present study, as follows:

1) Research problem #1: the heat distribution of the new self-designed,

custom-made prototype LED luminaires, which adopt the new system

architecture of “Heat Arrangement of LED Arrays in Low Profile”, is

unknown and not yet specified in any existing codes;

2) Research problem #2: the heat distribution pattern, as a desired thermal

parameter of the prototype LED fixtures, cannot be directly input into

Energy Plus for the annual heating and cooling energy simulation in that

Energy Plus is incapable of simulation of new LED technologies not yet

available in the program’s library which has a list of off-the-shelf

products available in the market.

To solve the first research problem (the heat distribution pattern of the new LED

luminaires is unknown and not yet specified in the existing codes), a portable calorimeter

was designed and constructed in house (Diederich, 2017) to measure the heat distribution

patterns of common lighting fixtures, including those LED luminaires used in the present

study. The second research problem (Energy Plus is incapable of simulation of new LED

technologies not yet available in the program’s library) could be solved via numerical

approximation in Energy Plus by assigning two separate “heat sources” in the room space

and ceiling plenum of the model classroom, respectively, coordinated with target lighting

schedule of the specified building space. In Energy Plus simulation, such a “heat source”

is only an approximation of the thermal characteristics of the tested LED luminaires

without considering the visible light fraction. As a result, in the present study, lighting

power of a LED luminaire will be ignored and the nominal power of the tested LED

luminaires consists only “heating power” of a luminaire, used for thermal related

simulation in Energy Plus via numerical approximation.

1.2 Research Goal and Objectives

The goal of the present study was to validate the energy-saving design of new LED

fixtures that deploy the new system architecture of “Heat Arrangement of LED Arrays in

Low Profile” to harvest the otherwise wasted heat generated by the same LEDs for an

overall reduction in building energy consumptions for heating and cooling.

To that end, a laboratory method was developed to measure a desired thermal

parameter of the LED luminaire: the ‘conditioned space/ceiling plenum split’. The

‘conditioned space/ceiling plenum split’ of the heat generated by each LED luminaire is

defined as the fraction of the lighting power converted to the lighting heat gain of the conditioned space and the fraction of the lighting power converted to the ceiling plenum’s lighting heat gain. Values of the ‘conditioned space/ceiling plenum split’ were obtained by testing each LED luminaire in the calorimeter under various testing conditions. Results were then used for estimation of annual energy consumptions in follow-up computer simulations in Energy Plus.

Next, a computer model of a primary school classroom (28’×36’) was built in which the new LED luminaires were installed for energy simulations in Energy Plus. The heat distribution pattern of each LED luminaire was approximated in Energy Plus in light of those laboratory values of ‘conditioned space/ceiling plenum split’. Energy Plus simulation was conducted to estimate how much annual cooling and heating energy can be saved by the new type of LED luminaires that deploy the new system architecture of “Heat

Arrangement of LED Arrays in Low Profile”. Simulation results obtained in test case of modeling equipped with the new LED luminaires were compared to those obtained in control cases of modeling equipped with off-the-shelf commercial LED luminaires.

In order to achieve the goal, there were three objectives to accomplish in the present study, as follows:

(1) Conduct a pilot study using experiments carried out in the calorimeter

put in the Dark Room (1152 Learned Hall) of the University of Kansas

Lighting Research Laboratory to explore some key information of the

calorimeter and the thermal experiments in preparation for formal

experiments.

(2) Conduct formal experiments using the calorimeter put in the Cold Room

(G445A M2SEC building) of the Engineering complex, which is a well-

controlled interior space with preset constant air temperature and

relative humidity, which are adjustable, to explore the thermal

performances (e.g., heating power and heat distribution pattern) of the

new LED luminaires under different test conditions in comparison to

those of selected commercial LED luminaires.

(3) Conduct Energy Plus simulation to calculate annual heating and cooling

energy consumptions of the modeled primary school classroom installed

with different LED luminaires previously tested in the Cold Room.

Additionally, the energy-saving modeling built in Energy Plus was used

to find out how much energy can be saved by utilizing the portion of

LED heat harvested in the room space as a supplemental reheat source

in cooling season.

According to the two unique properties of LED chips aforementioned in chapter

1.1, LED luminaires have heat distribution that is more easily controllable than other types of obsolete luminaires (e.g., fluorescent, HID luminaires). The new system architecture of

“Heat Arrangement of LED Arrays in Low Profile” was used in the present study to harvest the majority of heat generated by the LED chips and dissipate the LED heat into the room space while minimizing the heat flow toward the back of luminaires that would be trapped in the ceiling plenum. Figure 9 shows the concept of integrative illumination and HVAC system, in which the harvested LED heat is dissipated into the conditioned space to interact

with the HVAC system for overall energy savings. As shown in Figure 9, LED heat is harvested from multiple types of the new LED luminaires, including ceiling mounted, pendant, table top, floor standing fixtures, and LED-integrated air diffusers. The majority of harvested LED heat is directly dissipated into the conditioned room space to reduce space heating loads in winter and the reheating power of HVAC system in summer. To maximize energy saving potentials, ideally all of the LED heat could be harvested in the room space with zero heat trapped in the ceiling plenum so that the heat distribution pattern

(‘conditioned space/ceiling plenum split’) of this new LED technology is 100/0 (100% in room space, 0% in ceiling plenum).

Figure 9 The concept of integrative illumination and HVAC system, in which the

harvested LED heat in conditioned space interact with the HVAC system (Cai, 2016)

Using the calorimeter in the Cold Room, this study tested several thermal parameters of the prototype LED luminaire (one prototype with two different mounting

methods) with the new system architecture of “Heat Arrangement of LED Arrays in Low

Profile” and, for comparison, two off-the-shelf LED fixtures bought from market. The prototype new LED fixtures have nominal power of 54W, as designed, with more heat dissipated down into the room space than going up and trapped in the ceiling plenum. In the Cold Room experiments, the heat generation rate of each type of LED luminaire and its heat distribution pattern (‘conditioned space/ceiling plenum split’) were calculated using some thermal equations validated in the pilot study.

The final part of the present study was Energy Plus simulation with input from the laboratory tests, including the heating power of each type of LED luminaire and its heat distribution pattern (‘conditioned space/ceiling plenum split’). Accordingly, the thermal features in the pre-simulation settings of Energy Plus modeling were defined, and modified later per need, to approximate thermal parameters of LED luminaires obtained in laboratory. In cooling season, the LED heat harvested in the conditioned room space was simulated in Energy Plus as a supplemental reheating source of the HVAC system. It is worth mentioning that during this simulation process, “schedule conflict” was a concern in that the space lighting schedule does not perfectly match the space heating and the cooling schedules. When Energy Plus simulations were done, the percentage of overall building energy saving for space heating and cooling could be estimated by comparing the energy consumption data among different thermal pre-settings related to different types of LED luminaires installed.

CHAPTER 2

LITERATURE REVIEW

2.1 Heat gains from space lighting

Lighting accounts for 10% of total building energy consumption in commercial

building in the United States (U.S Energy Information Administration, 2015). Since all

electric lights generate heat, which is considered one of the major heat gains of buildings

with significant thermal impact on increased building cooling load, if not efficiently

removed out of the building. ASHRAE (American Society of Heating, Refrigerating, and

Air-conditioning Engineer) recommends a procedure to calculate the cooling load

incurred, which uses the ‘conditioned space/ceiling plenum split’ as an input of heat

distribution from luminaires (ASHRAE RP-1282, 2016).

Different types of heat sources generate different proportions of heat. For

incandescent lamps, often 80% of input electric power is transferred to thermal radiant,

10% converted to convective and conducted heat, and only 10% transferred to visible light

for space lighting (Carrier Air Conditioning Company, 2017);

U.S Department of Energy provide typical heat conversion for “white” light source

as shown in Figure 10 (U.S. DoE, Mar 2008). The LED light source have negligible

fraction of IR and UV and most of heat (conducted and convected) is dissipated into

ambient space. The ceiling plenum/conditioned space split of these 70%-80% conducted

and convected heat of LED luminaires that used for this study will be measured by the

calorimeter.

ASHRAE also gives some recommended split of sensible heat gain

(radiant/convective) from space lighting with different types of light sources such as incandescent, fluorescent, metal halide, and LED. (Engineer Reference, Big Ladder

Software LLC, 2018).

Figure 10 Heat conversion for typical light sources (U.S.-DoE, 2008)

Heat gain from a lighting fixtures often consists of three fractions: (i) converted heat in the room space, (ii) radiant heat into the room space, and (iii) wasted heat in the ceiling plenum hauled away by zone return air, which is commonly seen in a building equipped with lighting fixtures integrated with return air duct (Figure 11) (Input and

Output Reference, 2017, bigladdersoftware.com). Often, three fractions of heat gain from lighting fixtures, including converted and radiant heat in the room space and the trapped heat in the ceiling plenum, are simulated in Energy Plus.

Fisher & Chantrasrisalai (2006) conducted several experiments on heat distributions of overhead fluorescent fixtures, including ceiling mounted and pendant fixtures, in their research “Lighting Heat Gain Distribution in Buildings” (ASHRAE 1282-

RP, 2007). Their results are summarized in Table 1. Some kinds of lighting fixtures are designed to attach the return air plenum (as shown in figure 11) or designed to integrated with return air duct. The return air fraction of some lighting fixture is the fraction of heat absorbed by return air.

Not like conventional fluorescent, which may not output 100% lighting power in cold ambient environment, LED lighting fixture can be installed at the supply air side

(55°F). Therefore, heat generated by LED fixture might be utilized as part of space heating power and/or part of reheating power. The fraction of heat going down to the room space and going up to ceiling space is named ceiling plenum/room split. The fraction of heat in the room space could be regarded as kind of “supply air fraction” in this study.

Figure 11 Heat distribution of a fluorescent fixture (bigladdersoftware.com)

Table 1 Fractions of selected overhead fluorescent luminaires (Fisher & Chantrasrisalai, 2007)

Type of Luminaires Return Air Fraction Fraction Fraction Fraction Radiant Visible Convection Recessed, Parabolic Louver, 0.31 0.22 0.20 0.27 Non-Vented, T8 Recessed, Acrylic Lens, Non- 0.56 0.12 0.20 0.12 Vented, T8 Recessed, Parabolic Louver, 0.28 0.19 0.20 0.33 Vented, T8 Downlights, Compact 0.86 0.04 0.10 0.00 Fluorescent, DTT Downlights, Incandescent, 0.29 0.10 0.6 0.01 A21 Pendant, Indirect, T5HO 0.00 0.32 0.25 0.43 Surface Mounted, T5HO 0.00 0.27 0.23 0.50

Likewise, Rock and Wolf (1997) conducted several computer simulations to find

relevance of the heat gain from space lighting with different air supply system. In their

study case#2, the air supply and return ducts were equipped in ceiling plenum to haul away

possible heat gain from the lighting fixtures by the supply air as it passes through the air

ducts. In their simulation results, the supply air temperature maintained at 15°C during the

time when HVAC system and lighting system were on (Rock B.& Wolf D., 1997). It was

found that the air temperature of the ceiling plenum dropped by about 1.5°C during the

time when both HVAC and space lighting system were on.

Ball & Green (1983) created a mathematical model in 1983 to calculate cooling

loads from space lighting for a variety of space lighting arrangements. Chung & Loveday

(1998) create a thermal model, which consider all heat transfer including conduction,

convection and radiation of luminaires to analyze and calculate interior light level, room

air temperature and cooling loads (Diederich, 2017). These two studies

Moreover, various methods have been developed to test heat output of luminaires

to find its ‘radiative/convective split’ (the fraction of the lighting heat gain of the

conditioned space that is transferred as radiation and the fraction that is transferred as

convection) and the ‘conditioned space/ceiling plenum split’ (the fraction of the lighting

power converted to the lighting heat gain of the conditioned space and the fraction of the

lighting power converted to the ceiling plenum’s lighting heat gain), and new cooling load

calculation methods recommend these two splits as input data (AHRAE-RP-1282, 2018).

Unfortunately, both ‘radiative/convective split’ and ‘conditioned space/ceiling plenum

split’ cannot be estimated through simulation in Energy Plus or another software with

reliable results, thus, relying on laboratory method (as measured in calorimeter) to figure

them out.

2.2 Calorimeter

Calorimeter is one type of instrument commonly used to measure heat of chemical

reaction and/or physical change. Mitalas (1973) and his research team of National

Research Council of Canada built a full-size room calorimeter to calculate cooling load

from space lighting. They designed an air-loop with the same temperature of the insulated

wall circulated behind the insulation layers to prevent heat loss. Chantrasrisalai & Fisher

(2007) built a room calorimeter in 2007 to calculate cooling load by using air temperature

difference between supply and return air. Also, they installed a radiometer to measure

‘radiative/convective split’ of luminaires. Their research was published in the ASHRAE

handbook chapter 18, which indicates how to calculate cooling loads by using heat balance

and the radiant time series method (Diederich, 2017).

To calculate the heating power and its ‘conditioned space/ceiling plenum split’ of

LED luminaires, Cai and Diedrich designed and built a room calorimeter with simulated

ceiling plenum cavity and room cavity in the University of Kansas Lighting Research

Laboratory. The theory (Cerbe & Wilhelms, 2011) was used as a reference for the design

of the calorimeter as well as output heat power calculation. Two sets of air intake and

exhaust systems were installed in the calorimeter in the ceiling plenum and room space,

respectively (Diederich, 2017). The room space or ceiling plenum had three temperature

sensors to detect real-time air temperature at the air intake, above tested luminaire and at

the air exhaust, respectively, as well as one humidity sensor mounted in the middle of the

cavity, one fan sensor at air exhaust and one controllable fan to suck air (Diederich, 2017).

Air temperature and relative humidity were recorded by HOBO U-30 station data logger,

and the air flow rate was recorded by Testo-480 data logger. Based on Cerbe & Wilhelms’s

theory, the heating power of LED luminaires was calculated using air temperature

difference between the air intake and the air exhaust (measured with temperature sensors)

and air-flow rate (measured with fan probes). Next, the heat distribution pattern was

calculated by comparing the heating power output from the “room space” and that from

the “ceiling plenum” of the calorimeter. This calorimeter was tested and commissioned by

Simon Diedrich in his Master thesis study in 2017.

2.3 Reheat system

Reheat system is a component of building air-conditioning system, which can be

found in large buildings such as hospital, offices, commercial buildings, etc. The main

benefit of reheat system is to control the interior air temperature and humidity at a

comfortable level. As shown in Figure 11, returned air from conditioned space will be cooled down to a temperature below dew point to remove moisture (called condensation) for humidity control purpose. The chilled air is then re-heated to a comfortable temperature range by the reheat system before it is discharged into the room space.

The major shortcoming of reheat system is its high energy consumption, which may result in an extra cost for building utilities, especially for a building equipped with the electric coil reheat system. (Madison Gas and Electric Company, 2015).

Figure 12 A typical reheat system in commercial building (MGE, 2015)

CHAPTER 3

METHODOLOGY

3.1 Theoretical framework

3.1.1 Calculations

(a) The heating power of tested LED luminaires

Diederich (2017) selected and verified the following calculations process with

related equations for data treatment in his thesis, which were used in the present study for

covering all mathematical bases for calculating calorimeter output values.

The total output heating power consists of heating output in the ceiling cavity and

that in the room cavity, respectively, as shown in Equation (1).

푄̇푡표푡푎푙 = 푄̇푐푒𝑖푙𝑖푛𝑔 + 푄̇푟표표푚 (1)

Where:

푄̇ total: total heating power of the tested luminaire

푄̇ ceiling: heating power in the ceiling plenum cavity

푄̇ room: heating power in the room cavity

For each cavity, the heating power is calculated with the air temperature difference

and the mass flow rate using Equation (2).

푄̇푐푎푣𝑖푡푦 = (ℎ표푢푡 − ℎ𝑖푛) ∙ 푚̇ (2)

Where:

푄̇ cavity: heating power output from room cavity or ceiling cavity

ℎ표푢푡: Specific enthalpy of air leaving the calorimeter

ℎ𝑖푛: Specific enthalpy of air entering the calorimeter

푚̇ : The mass flow rate of air (kg/s)

The specific enthalpy of air is the function of air temperature and absolute humidity at this specific temperature, calculated using Equation (3) (Cerbe & Wilhelms, 2011). In

Equation (3), air temperature is measured with temperature sensors in the present study.

ℎ = 퐶푝퐿 ∙ 푡 + 푥 ∙ (퐶푝퐷 ∙ 푡 + 훾퐷) (3)

Where:

퐶푝퐿: Specific heat capacity of air, 1.004 KJ/kg*K

퐶푝퐷: Specific heat capacity of steam, 1.86 KJ/kg*K

t: Air temperature (Kelvin)

훾퐷: Evaporation heat at 273.15K, 2500KJ/kg

푥: Absolute humidity (푘𝑔푤푎푡푒푟/푘𝑔푑푟푦 푎𝑖푟)

The absolute humidity is calculated using Equation (4). In Equation (4), water vapor saturation pressure at specified temperature can be calculated using online engineering toolbox (www.engineeringtoolbox.com), and that temperature shall be the laboratory environment air temperature in the present study. Barometric pressure on the test date can be obtained online from weather.com.

푝푠(푡) 푥 = 0.622 ∙ 푝 (4) −푝 휑 푠(푡)

Where:

푝푠(푡): Water vapor saturation pressure (kPa)

p: Barometric pressure (kPa)

휑: Relative humidity (range: 0-1)

The mass-flow rate of air can be calculated from air volume-flow rate and the absolute humidity, using Equation (5) (Cerbe &Wilhelms, 2011). The volume-flow rate of air can be calculated using the air-flow rate (measured by TESTO vans probe) and the diameter of outlet at the air exhaust side (obtained from the specification sheet of the

TESTO vans probe).

푣̇ ∙푝 1 푚̇ = ∙ (5) 푅퐷∙푡 (0.622+푥)

Where:

푣̇: Volume flow rate of air (푚3/s)*

푝: Barometric pressure (Pa)

푅퐷: Specific gas constant water vapor, 461 J/kg*K

푡: Air temperature (K)

푥: Absolute humidity (푘𝑔푤푎푡푒푟/푘𝑔푑푟푦 푎𝑖푟)

푚̇ : The mass flow rate of air (kg/s)

(b) The heat distribution pattern (‘conditioned space/ceiling plenum split’)

Equation (6) is used to calculate the ‘conditioned space/ceiling plenum split’.

푄̇푟표표푚 푅푟표표푚/푐푒𝑖푙𝑖푛𝑔 = (6) 푄̇ 푐푒푖푙푖푛푔

Where:

푅푟표표푚/푐푒𝑖푙𝑖푛𝑔: The heating power split of tested luminaires

푄̇ ceiling: heating power in the ceiling plenum cavity

푄̇ room: heating power in the room cavity

In Energy plus simulations, the space heating energy (in heating season) and the reheating energy of the air-conditioning system (in cooling season) are automatically counted as one output value for annual heating energy consumption. Therefore, the reheating energy can be extracted by the following process in Equation (7).

푄푅푒ℎ = 푄ℎ,푎푛푛푢푎푙 − 푄ℎ푒푎푡𝑖푛𝑔 표푛푙푦 (7)

Where:

푄푅푒ℎ: Reheating energy generated in cooling season (GJ)

푄ℎ푒푎푡𝑖푛𝑔 표푛푙푦: Heating energy only for space heating (GJ)*

푄ℎ,푎푛푛푢푎푙: Annual heating energy (GJ)

It is worth mentioning that the 푄ℎ푒푎푡𝑖푛𝑔 표푛푙푦 can be obtained from the annual heating energy results of Energy Plus simulation of the tested building modeling containing only heating equipment (e.g., electric coil, gas heating, hot water heating) in its HVAC system (which means the cooling equipment and reheat system shall be removed before running simulation).

3.1.2 Research Assumptions

Due to some limitations existed in the Cold Room experiments and Energy Plus simulations, as detailed below, the present study was conducted based on the following assumptions:

Assumption #1: Heat gain from LED driver is not counted in the calculation.

The prototype LED luminaire with new architecture of “Heat Arrangement of LED

Arrays in Low Profile” developed in KU Lighting Research Lab was used to verify the idea that the new architecture of LEDs can redirect heat flux toward the front to utilize more heat generated by LED chips in the room space. Since the prototype LED was an incomplete product without an attached LED driver (instead, a DC power source was used), there was no way to count heat gains from LED driver into the total heating power of prototype LED luminaires. Therefore, for prototype LED fixtures tested in this study, we assumed no heat gain from LED drivers was counted in the calculation. To be consistent,

LED drivers and/or DC powers connected to all other luminaires tested in this study were put outside of the calorimeter to power LEDs inside.

Assumption #2: Using the calculated heating power of the LED luminaires as the input values of lighting load (w/푚2) in Energy Plus simulation.

Based on the design of the calorimeter, heat gain in each cavity from the tested lighting fixtures was calculated using air temperature difference measured with air

temperature sensors. Using available calorimeter’s output values, it was impossible to calculate the ‘radiative/convective split’ of tested LED luminaires in the calorimeter.

Fortunately, the heat contribution from radiation to the temperature measurement in calorimeter was negligible because LEDs have almost zero IR radiation in the forward direction (different from conventional light sources). Also, the inside surfaces of the calorimeter were covered with aluminum foils with low emissivity. The temperature sensors also had a very tiny surface area and made of low emittance materials (Figure 12).

Since the focus of this study was the thermal performance of LED fixtures, the lighting power density (W/푚2) in the Energy Plus would only contain pure heating power generated from luminaires with the fraction of visible light (in a range 0-1), while the fraction of radiant (in a range 0-1) were set to 0 (Figure 13).

Figure 13 Two types of sensor installed in the calorimeter with aluminum foil covered

interior surfaces

Figure 14 Lights definitions setting page in Energy Plus

Assumption #3: In Energy Plus simulation for LEDs with new lighting arrangement, in cooling season, the heating power of LED luminaire in room cavity is assumed to be 0, which is otherwise taken as increased cooling load by the Energy Plus software, to realize the energy saving design that utilizes the harvested room-portion of heat generated by LEDs as a supplemental reheating power of the building air-conditioning system.

The LED heat generated by the new LED luminaires and harvested in the room space can be used to supplement space heating in heating season, which is relatively easy to simulate in Energy Plus. Nonetheless, in cooling season, the heat gain from the conventional lighting system often increases the cooling load of building HVAC system.

To change this situation, the installation and operation of future solid-state lighting system should be integrated with the building HVAC system. The innovation of the new

integrative LED luminaires is to harvest the otherwise wasted heat generated by the LEDs and then utilize the heating power of luminaires in the room space to warm up the chilled air as a supplemental re-heating power to the reheating system of the building air conditioning system in cooling season. Therefore, when this integrative illumination and air-conditioning system is applied in buildings, the heating power of LED luminaires

(calculated from the laboratory experiments results) harvested in the room space shall be considered as a supplemental reheating power instead of an undesirable cooling load in cooling season. Accordingly, in Energy Plus simulation, the heating power of LED luminaire in room cavity shall be set to 0, which is otherwise taken as increased cooling load by the Energy Plus software, to realize the energy saving design that utilizes the room- portion of heat generated by LEDs as a supplemental reheating power in the cooling season.

Assumption #4: In Energy Plus simulation, all four luminaires tested in the laboratory have approximately the same heating power.

Different LED luminaires often have different thermal features (e.g., heating power, heat distribution), resulted from different system architectures, different materials, different LED chips and their layout on the MCPCB board, and different insulation, etc.

Thus, even with the same nominal wattage, the four LED luminaires tested in the laboratory may not have the same value of heating power harvested in the room space.

One goal of the present study was to validate that utilizing the room-portion of heat generated by LED luminaires that is harvested in the room space can save annual heating

and cooling energy of buildings. To that end, the annual heating and cooling energy

consumptions of different LED luminaires installed in the primary school classroom were

simulated in Energy Plus and compared. The primary variable in Energy plus simulation

shall be the heat distribution pattern (‘conditioned space/ceiling plenum split’) of those

luminaires. All other variables, including the heating power of each LED luminaire, may

have mixed effects on the simulation results, thus, shall be excluded by assigning preset

values to those confounding variables. Therefore, the total heating power (room portion

plus ceiling portion) is set to 50W for all luminaires used in Energy Plus simulation.

3.2 Pilot study

A pilot study was conducted to collect useful information in preparation for the

formal experiments, such as potential issues that may affect the formal experiments, which

need to be identified and resolved in the pilot study. Meanwhile, all sensors were checked

to ensure they would all be under normal working condition in the formal experiments.

The calculation process for data treatment was also verified in the pilot study.

3.2.1 “Dark-room” experiments

The pilot experiment was conducted in the Dark Room (1152 Learned Hall) to

explore key factors that may have essential influence on experiment results and/or output

values. A total of ten tests were conducted in the Dark Room, as shown in Table 2,

including five tests of surface-mounted LED luminaires and another five tests of ceiling-

recessed LED luminaires. Figure 15 shows the prototype LED with two typical

constructions: ceiling recessed (15-a) and surface mounted (15-b) in this study. Figures 16-

19 are the actual luminaires testing process in the Dark Room under room temperature.

Table 2 Configurations of LED luminaires’ tests in pilot study

Test case # Luminaire type Mounting method Nominal Power 1 Prototype LED Ceiling Recessed 57.6W 2 Prototype LED Ceiling Recessed 57.6W 3 Prototype LED Ceiling Recessed 57.6W 4 Prototype LED Ceiling Recessed 57.6W 5 Prototype LED Surface Mounted 57.6W 6 Prototype LED Surface Mounted 57.6W 7 Prototype LED Surface Mounted 57.6W 8 Prototype LED Surface Mounted 57.6W

(a) (b)

Figure 15 The same prototype LED luminaires tested in the calorimeter were mounted

differently in different tests, (a) ceiling-recessed, (b) surface-mounted

Figure 16 The prototype LED luminaire surface mounted on the ceiling of the calorimeter

chamber

Figure 17 Data recording during tests

Figure 18 Temperature sensor mounted at air intake of the room cavity

Figure 19 The ceiling recessed prototype LED luminaire in the calorimeter

3.2.2 Data treatment and results

Table 4 shows the calculation results of heat output of LED luminaires from the

calorimeter with trusted input values. These input values are either direct measured by

calorimeter (e.g. temperature difference between air-intake and air-exhaust, relative

humidity inside of the calorimeter, air-flow rate) or obtained from authority agency (e.g.

Climate data from local weather station website). Barometric pressure (Table 3) was

downloaded from weather station website (weather.com). Other values measured inside

the calorimeter, including relative humidity (%), air-flow rate (m/s) and air temperature

(°F/°C), were readout after heat balance was reached inside the calorimeter. Water vapor

saturation pressure (Figure 19) was obtained from water saturation pressure chart.

Table 3 Barometric pressure (Daily weather report, available at weather.com, 2018)

Barometric Pressure, Lawrence, KS Date In Hg kPa 08/21/18 30.12 101.9900 08/22/18 30.30 102.6079 08/27/18 29.80 100.9147 08/28/18 29.80 100.9147 08/29/18 30.10 101.9306 09/05/18 30.10 101.9306 09/07/18 30.10 101.9306 09/10/18 30.00 101.5920 09/12/18 30.00 101.5920 09/14/18 30.00 101.5920 09/28/18 30.10 101.9306 09/27/18 30.10 101.9306 10/01/18 30.00 101.5920 10/04/18 30.10 101.9306 10/05/18 30.10 101.9306 10/03/18 30.00 101.5920 10/08/18 30.00 101.5920 10/10/18 30.10 101.9306

Figure 20 Water vapor saturation pressure chart [Engineering ToolBox, (2001). Available at:

https://www.engineeringtoolbox.com, Dec 2018]

Table 4 lists all the test results from the pilot study experiments with different

configurations (Table 2). There are three issues found in Table 4, which need further

discussions. Firstly, the ratio of the wattage loss of the calorimeter (due to thermal leakage

through imperfect insulation, the output wattage is less than input wattage) to the total input

power is not a constant value, even for the same luminaires under same testing

configurations (as shown in Test 1 and Test 2). Secondly, the ‘conditioned space/ceiling

plenum split’ of the same LED luminaires was not a constant value when obtained from

different tests in the calorimeter under the same configuration of experiment (e.g., with the

same mounting method and the same input power, as in Test 5 and Test 7). Thirdly, in Test

5, the wattage loss was a negative value, which is incorrect since the output heating power

cannot exceed the input heating power according to the first law of thermodynamics. Those

three issues observed in the pilot study were caused by the uncontrolled varying testing

environment (e.g., room temperature and humidity) in the Dark Room. During the process

of the experiments conducted in the Dark Room, the AC unit in the Dark Room kicked in

from time to time resulting in fluctuating room temperature. The random occupancy of

other users who accessed the Dark Room during the experiments also resulted in unknown

impact on the room temperature, humidity and air circulation that would affect the ongoing

experiments. Those uncertainties could be avoided in a well-controlled test environment.

All those three issues were overcome in the follow-up formal experiment carried out in the

well-controlled space, “Cold Room”.

Table 4 Results from the “dark room” experiments

Test # Test 1 Test 2 Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Date 08/21/18 08/22/18 08/27/18 08/28/18 08/29/18 09/05/18 09/07/18 09/10/18 Input power 26.30 26.30 30.60 54.0 54.0 63.0 54.0 54.0 (W) Total output 24.90 21.39 26.74 54.43 49.78 55.97 46.11 45.92 power (W) Power in room 14.23 12.22 15.06 27.21 23.91 33.81 29.99 29.40 (W) Power in 10.67 9.17 11.68 27.21 25.88 22.16 16.11 16.52 ceiling (W) Wattage loss 1.40 4.91 3.86 -0.43 4.22 7.03 7.90 8.08 (W) Wattage 5.3% 18.7% 12.6% -0.8% 7.8% 11.2% 14.6% 14.9% loss/total (%) “room/ceiling” 57/43 57/43 56/44 50/50 48/52 60/40 65/35 64/36 split

3.2.3 Recommended improvements for formal experiment

The Dark Room was not a well-controlled space for maintaining constant environmental temperature and humidity over time. During the tests, the room temperature and humidity were varying because of the unpredictable working condition of the independent AC unit, resulting in unstable temperature balance point of the calorimeter

(the temperature difference between air-intake and air- exhaust is no longer changing), which largely affected the accuracy of measurement results for data treatment.

Following are three recommendations to solve those three issues observed in the pilot study for improvements in the follow-up formal experiment to be conducted in the

Cold Room.

(a) Each experiment should be conducted in a closed, well-controlled and

undisturbed space.

(b) To reach the heat balance of the calorimeter, each experiment should be

lasted for at least 5 hours.

with aluminum foil, making it the same construction for minimum heat

radiation as those commercial LED luminaires with aluminum housing.

Since the Cold Room installed with special air-conditioning equipment is a well- controlled space with constant room temperature and relative humidity, which are adjustable, those three issues aforementioned in the pilot study would not be observed in the Cold Room experiments. Additionally, the formal experiments in Cold Room would

not be disturbed to prevent external interference on the stable room temperature and the

relative humidity. Each experiment would last at least 5 hours to ensure that the calorimeter

reaches heat balance point (ongoing heat generated by tested LED luminaires is quickly

removed by controlled air-flow that reaches an equivalence status).

3.3 Experiments in the Cold Room

3.3.1 Introduction

The Cold Room (G455A) in M2SEC building is a well-insulated laboratory with

an interior environment control system providing stable room temperature and relative

humidity, which are preset values. The interior air temperature of the Cold Room can be

set from 0°C to 30°C with an accuracy of one decimal and relative humidity inside could

be keep at range 45%-50%. Nonetheless, it is worth mentioning that the environment

control system stabilizes interior air temperature and relative humidity by using strong

circulating air-flow, which may increase the heat loss of the calorimeter due to increased

convection occurred on exterior surface of the calorimeter.

Figure 20 shows the Cold Room laboratory and the outside control panel that enable

adjustment of the interior air temperature without opening the door causing undesired

external interference on the ongoing experiment. Figure 21 shows the setup of the

experiments in the Cold Room. All air pipes of the calorimeter were connected to fans put

on an equipment cart. Temperature and humidity sensors in the calorimeter were connected

to HUBO data logger and air-flow sensors connected to TESTO-480 data logger put on the

equipment cart.

(a) (b)

Figure 21 The Cold Room laboratory, (a) the control panel, (b) a watch window

on the door of the Cold Room

Figure 22 Setup of the calorimeter in the Cold Room

To keep an isolated environment in the Cold Room, the door was closed for ongoing experiments to avoid any unnecessary external interference until a set of experiments were completed without need of changing fixtures or mounting methods (e.g. testing surface- mounted custom-designed LED luminaire under four different air temperatures). The DC power, patch panel and data-loggers, were locked inside of the Cold Room connected to the electricity outlet and laptop (for data recording) through a dedicated wire hole insulated with insulation materials to outside of the “cold room” (Figure 23). The ongoing experiment inside the Cold Room was observed through a double-glazing window with anti-condensation design (Figure 24).

Figure 23 Power cable and data wire

Figure 24 Anti-condensation observation window

3.3.2 Calibrations

Two calibrations were conducted before testing LED luminaires in the Cold Room.

First, the room temperature of the Cold Room was measured by three temperature sensors and checked with the interior temperature value shown on the control panel. When the two readings did not agree, the reading of temperature sensors were calibrated to that of the control panel.

As shown in Figure 24, the calibration was conducted at a reasonable range of the ambient temperature from 4°C to 30°C with a step of 2°C increase. It was found that the data curves of three sensors (one mounted at the top air-intake for measuring the air temperature of the ceiling cavity of the calorimeter, one at the down air-intake for measuring the air temperature of the room cavity, the third one mounted outside of the calorimeter in the Cold Room for measuring the cold room temperature) were very close to each other at the same trend of changes. Their read-out temperatures (in °C) were also very close to the room temperature values read from the Cold Room control panel.

However, as shown in Figure 24, the data curves of other six temperature sensors inside the calorimeter showed a “lagging pattern” behind the changes of the room temperature of the Cold Room, indicating the air temperature inside the two Calorimeter cavities did not change immediately with changes of the room temperature of the Cold

Room, caused by thermal resistance of the calorimeter. To solve this problem, longer testing time was necessary for components inside the calorimeter reaching thermal equilibrium to reduce the lagging pattern caused by thermal mass of the calorimeter. Thus, a second experiment was conducted with increased room temperature of the Cold Room from 12.5°C to 21°C (step 1) and then from 21°C to 29.5°C (step 2) to test the temperature

sensors read-out from one temperature balance state to another one. At each step, after the

Cold Room temperature was adjusted to the new value (21°C at step 1, 29.5°C at step 2),

the calorimeter was allowed for 6 hours to reach its temperature balance to solve the

“lagging” issue of the six temperature sensors inside. As shown in Figure 25, it was found

that all sensors were working normally according to their read-out temperatures which

were the same value when the thermal balance point was reached.

*Format of sensors: sensor number-sensor’s type- (sensor’s location)

Figure 25 Results of the first calibration: temperature increased from 4°C to 30°C (39°F to

86°F) with each step is 2°C

Figure 26 Calibration of temperature sensors with tested room temperature of 21.0°C to 29.5°C

(70°F to 85°F).

In addition to the first calibrations of the temperature sensors mounted inside the

Calorimeter, a second calibration was conducted for compensation of the thermal loss of

the calorimeter. Diederich (2017) already conducted several experiments in 2017 to

determined heat-loss rate of the calorimeter. In his experiments, a cartridge heater (pure

heat source with 100% input power transferred to heat, as shown in Figure 27) was mounted

inside the calorimeter to obtain heat loss of the calorimeter. Based on previous studies of

the commissioning of the calorimeter, 11% of wattage loss is defined as an acceptable

measurement error (Diederich, 2017). A similar experiment was conducted in 2018 to

repeat Diederich’s calibration procedure to make sure the calorimeter was well-insulated

and still in good working condition before the formal experiment. Figure 28 shows the

temperature readings over the calibration process of the calorimeter heat loss over six

hours. As shown in Table 5, the input power of the cartridge heater was 35.31 W (recorded

from the DC power) while the calculated heat output was 32.61 W based on the measured

values of the calorimeter after thermal equilibrium was reached. Therefore, the heat loss

rate was calculated as 7.6% in this calibration test, as shown in Table 5, which was less

than 11% acceptance level set by Diederich’s experiment, showing the calorimeter was

still in normal working condition ready for formal experiments.

Table 5 Results of the cartridge heater tests in the Calorimeter

Calibration test-cartridge heater reference value Measured value Room Cavity N.A. 12.27 W Celling Cavity N.A. 20.34 W Total Output N.A. 32.61 W INPUT U=32 V, I= 1.1 A 35.3 W ROOM T 20.0 °C (68 °F) 20.38 °C (68.7°F) Wattage loss N.A. 2.69 W Heat loss ratio N.A. LOSS/T=7.6%

Figure 27 The pure heat source (cartridge heater) mounted in the Calorimeter

Figure 28 The Calibration of calorimeter heat loss at room temperature of 20°C, showing the

calorimeter reached thermal equilibrium over 6 hours

With the three calibrations completed, the calculated heat output from the LED

luminaire tested in the calorimeter could be trusted values. Temperatures measured with

those sensors had a satisfied accuracy of within ±0.5°F , which reflects the difference

between the temperature measured by the sensors mounted inside and outside of the

calorimeter and the preset value on the “cold room” control panel. The heat loss of the

calorimeter was 7.2% (cartridge heater at 35W input), which is in the tolerance range

(within ±10%) based on previous study by Simon Diederich in 2017. As a result, the

calorimeter was considered in good working condition and ready for the formal

experiments in the Cold Room.

3.3.3 Formal experiments and the results

The purpose of the Cold Room experiments was to measure the heat distribution pattern (‘conditioned space/ceiling plenum split’) of the tested LED luminaires with two different mounting methods (ceiling recessed versus ceiling surface mounted). Table 6 shows a complete lists of eight tests in this experiment, including ceiling-recessed prototype LED fixture tested in summer and winter conditions (Tests 1 and 2), ceiling surface-mounted prototype LED fixture in summer and winter conditions (Tests 3 and 4), and two commercial LED fixtures either ceiling-recessed (Tests 5 and 6) or surface- mounted (Tests 7 and 8) tested in summer and winter conditions. Each test contained four separate trials under four different room temperatures: 68°F , 72°F , 74°F , 78°F (ASHRAE standard 55 2010). As summarized in Table 7, the Cold Room temperature was set to 68°F and 72°F (ASHRAE recommended lowest and highest acceptable indoor temperatures in winter) to simulate luminaires working in the winter room’s conditions, while 74°F and

78°F in summer room’s conditions (ASHRAE recommended). During the experiment, to exclude external interference on the tests, the Cold Room was locked until one set of experiment was completely done under different room temperatures without necessary changes of luminaires and their mounting methods.

Table 6 “Cold room” luminaire tests configurations

Test # Tested Mounting Nominal Season for Acceptable Acceptable Luminaires method Power Simulated Room Temperature Temperature (Labeled) Condition Lower limit Upper limit

Test 1 Prototype Ceiling 57.6 W Summer 20 °C (68 °F) 22.2 °C (72 °F) LED Recessed Test 2 Prototype Ceiling 57.6 W Winter 23.3°C(74°F) 25.6°C(78°F) LED Recessed Test 3 Prototype Surface 57.6 W Summer 20 °C (68 °F) 22.2 °C (72 °F) LED Mounted Test 4 Prototype Surface 57.6 W Winter 23.3°C(74°F) 25.6°C(78°F) LED Mounted Test 5 Commercial Ceiling 50 W Summer 20 °C (68 °F) 22.2 °C (72 °F) LED Recessed Test 6 Commercial Ceiling 50 W Winter 23.3°C(74°F) 25.6°C(78°F) LED Recessed Test 7 Commercial Surface 2*25 W Summer 20 °C (68 °F) 22.2 °C (72 °F) LED Mounted Test 8 Commercial Surface 2*25 W Winter 23.3°C(74°F) 25.6°C(78°F) LED Mounted

Table 7 Guidelines of acceptable room temperature, excerpts (Thermal environment Condition

for Human Occupancy, ASHRAE Standard 55, 2010)

Type of space In Fahrenheit (°F) In Celsius (°C) Summer Winter Summer Winter residences, apartments 74-78 68-72 23-26 20-22 classroom, courtroom 74-78 68-72 23-26 20-22 school dining room 75-78 65-70 24-26 18-21 ballrooms 70-72 65-70 21-22 18-21 retails shop, supermarket 74-80 65-68 23-27 18-20 medical operating rooms 68-76 68-76 20-24 20-24 toilet room, service room 80 68 27 20 locker room 75-78 65-68 24-26 18-20

In addition, the controlled air-flow rates of both ceiling and room cavity were

calibrated to approximately the same value before a new set of experiment was started.

Figure 28 shows the real time measurement of the controlled air-flow. The data curve in

red is for ceiling cavity and that in purple is for room cavity.

Figure 29 The real time measurement of the controlled air-flow rate in

ceiling cavity (red) and room cavity (purple).

Tables 8-11 are the summarized results of eight luminaire experiments conducted in the Cold Room. Each table showing one luminaire with four typical room temperature

(upper limit and lower limit of acceptable room temperature in summer and winter). A complete list of outcomes with a detailed calculation process was available in Appendix 1

“calculation sheet of cold room luminaires’ tests”.

Based on the results shown in Table 8, the heat distribution pattern (‘conditioned space/ceiling plenum split’) of the ceiling-recessed prototype LED luminaire was determined as of 60.4/39.6, the average of similar values obtained in the four tests.

Similarly, the ‘conditioned space/ceiling plenum split’ was averagely 69.6/30.4 for the ceiling surface-mounted prototype LED luminaire (Table 9), 47.2/52.8 for the ceiling-

recessed commercial LED luminaire (Table 10), and 58.1/41.9 for the surface mounted

commercial LED luminaire (Table 11).

Table 8 Results of the ceiling recessed prototype LED luminaire testing

TEST # 1-a 1-b 2-a 2-b Room Temperature 20.0 °C (68 °F ) 22.22°C (72°F) 23.33°C (74°F) 25.56°C (78°F) °C (°F) Measured Room Temperature or Temperature of Air intake 20.28°C (68.5°F) 22.56°C (72.6°F) 23.61°C (74.5°F) 25.83°C (78.5°F) °C (°F) Air-flow rate 1.78 1.78 1.78 1.78 (m/s) Temperature Difference in room cavity(°C) 1.85 1.90 1.87 1.88 Temperature Difference in ceiling cavity(°C) 1.24 1.23 1.21 1.24 Measured heating Power in room cavity (W) 30.40 31.23 30.79 30.91 Measured heating Power in ceiling cavity (W) 20.38 20.21 19.94 20.42 Conditioned space/ceiling plenum split in each test 59.9/40.1 60.7/39.3 60.7/39.3 60.2/39.8

Table 9 Results of the surface mounted prototype LED luminaire testing

TEST # 3-a 3-b 4-a 4-b Room Temperature 20.0 °C (68 °F) 22.22°C(72°F) 23.33°C(74°F) 25.56°C(78°F) °C (°F) Measured Room Temperature or Temperature 20.28°C (68.5°F) 22.42°C (72.3°F ) 23.61°C (74.5°F) 26.04°C (78.8°F ) of Air intake °C (°F) Air-flow rate 1.78 1.78 1.78 1.78 (m/s) Temperature Difference in room cavity(°C) 2.18 2.17 2.13 2.16 Temperature Difference in ceiling cavity(°C) 0.95 0.97 0.92 0.93 Measured heating Power in room cavity (W) 35.72 35.64 35.98 35.57 Measured heating Power in ceiling cavity (W) 15.62 15.93 15.09 15.30 Conditioned Space/Ceiling Plenum split 69.6/30.4 69.1/30.9 69.9/30.1 69.9/30.1

Table 10 Results of the ceiling recessed commercial LED luminaire testing

TEST # 5-a 5-b 6-a 6-b

Room Temperature °C (°F) 20.0 °C (68 °F) 22.22°C(72°F) 23.33°C(74°F) 25.56°C(78°F) Measured Room Temperature or 20.44°C (68.8°F) 22.67°C (72.8°F ) 23.66°C (74.6°F ) 25.74°C (78.3°F ) Temperature of Air intake °C (°F) Air-flow rate 1.68 1.68 1.68 1.68 (m/s) Temperature Difference in room cavity(°C) 1.54 1.61 1.67 1.52 Temperature Difference in ceiling cavity(°C) 1.73 1.74 1.88 1.74 Measured heating Power in room cavity (W) 23.87 24.84 25.85 23.78 Measured heating Power in ceiling cavity (W) 26.81 26.85 29.10 27.22 Conditioned Space/Ceiling Plenum split 47.1/52.9 48.1/51.9 47.0/53.0 46.6/53.4

Table 11 Results of the surface mounted commercial LED luminaire testing

TEST # 7-a 7-b 8-a 8-b

Room Temperature 20.0 °C (68 °F) 22.22°C(72°F) 23.33°C(74°F) 25.56°C(78°F) °C (°F) Measured Room Temperature or 20.44°C (68.8°F) 22.67°C (72.8°F) 23.78°C (74.8°F) 25.89°C (78.6°F) Temperature of Air intake °C (°F) Air-flow rate 1.65 1.65 1.65 1.65 (m/s) Temperature Difference in room cavity(°C) 2.38 2.33 2.22 2.32 Temperature Difference in ceiling cavity(°C) 1.73 1.66 1.65 1.62 Measured heating Power in room cavity (W) 36.59 36.90 34.37 35.91 Measured heating Power in ceiling cavity (W) 26.60 25.58 25.55 25.08 Conditioned Space/Ceiling Plenum split 57.7/42.3 58.4/41.6 57.3/42.6 58.9/41.1

3.4 Computer-Aided simulation in Energy Plus

3.4.1 Design of Energy Plus simulation

A DOE’s standard primary classroom equipped with those test LED luminaires was

used for computer-aided simulation of the annual energy consumption. The model

classroom has single floor, flat roof, located at Lawrence, KS, in ASHRAE climate zone

4. As shown in Figure 29, the room dimensions are approximately 11 m / 36 ft (length) by

8.5 m / 28 ft (width) by 3.05 m / 10 ft (height), with a ceiling plenum height of 0.91 m /3

ft. The model classroom was built in Sketch Up with Open Studio plugin and then input

into Energy Plus. Constructions of the classroom was defined by Open Studio based on

space type (primary school) and climate zone (4A). The general lighting was provided

using 15 LED luminaires with nominal power of 50W/each. The calculated lighting power

density (LPD) was 8.6 W/m2 based on the gross room area (93.5 m2) to be complied with

the requirement of lighting codes for primary school classroom (IES handbook 10th edition,

2011). Weather data were obtained from the weather station of Topeka-Forbes AFB

(energyplus.net/weather), which is the nearest station to Lawrence.

The present study focused on investigation of the thermal performance of LED

luminaires using the calculated heating power of the LED luminaires as the input values of

lighting load (w/m2) in Energy Plus simulation (Assumption #2). Thus, the light energy

fraction of the tested LED luminaires was not counted in Energy Plus simulation. The heat

fractions of the tested LED luminaires were approximated in Energy Plus by “pure heat

sources” (still considered a definition of lighting fixtures in Energy Plus). The replacement

“heat source” was then divided into two separate parts, with one assigned in the ceiling

plenum and the other assigned in the room space to approximate the heat distribution

pattern (‘conditioned space/ceiling plenum split’) of the tested LED luminaires obtained

from the prior calorimeter experiments in the Cold Room. Figure 30 illustrate how to input

the ‘conditioned space/ceiling plenum split’ of the tested LED luminaires in Energy Plus

simulation. Figure 31 shows how to set the heating power of tested LED luminaires in the

ceiling plenum space. In this simulation, all lighting fixtures were controlled on a typical

lighting schedule of primary school classroom (Figure 29).

Figure 30 Floor plan and lights layout of the primary school classroom (Cai, 2015)

Figure 31 Define the heat distribution of luminaire

Figure 32 Define the heating power of luminaire in ceiling plenum

In Energy Plus simulation, the packed rooftop VAV (Variable Air Volume) with PFP

(Parallel Fan Powered) boxes and Reheat was selected for the model classroom as its HVAC system with default heating and cooling schedule (through temperature set point) determined by the climate zone of the project site in Energy Plus, as shown in Figure 32.

Figure 33 Defining the HVAC system of the model classroom in Energy Plus

3.4.2 Introduction to the modeling with the integrated lighting arrangement design

The custom-made prototype LED luminaire was used to verify a new energy-saving design in which the harvested LED heat in the room space is used as a supplemental reheating power to the re-heating coil of the HVAC system. The energy-saving design uses an integrative solid-state lighting & heating arrangement called “Heat Arrangement of LED

Arrays in Low Profile”. The energy-saving feature of the custom-made LED luminaire with “Heat Arrangement of LED Arrays in Low Profile” is to utilize the otherwise wasted portion of heat generated from prototype LED luminaires and trapped in the room space as part of reheat source. Therefore, assumption #3 was taken in Energy Plus simulation (in cooling season, the heating power of LED luminaire in room cavity is assumed to be 0, which is otherwise taken as increased cooling load by the Energy Plus software, to realize the energy saving design that utilizes the harvested room-portion of heat generated by

LEDs as a supplemental reheating power of the building air-conditioning system). During

the cooling season, heat generated from luminaires and trapped in the room space is

considered as the supplemental reheating energy that shall be deducted from the original

reheating energy consumption with the only heat source of electric heating coil.

To approximate this energy-saving design in Energy Plus, the model classroom was

added with two updated features. First, the harvested heat of the tested LED luminaires,

which was otherwise trapped in the room space, was now removed. Second, the removed

heat was counted into the reheating energy of the HVAC system. As a result, in Energy

Plus simulation, both cooling load and reheating energy consumption could be reduced.

Figure 33 is an illustration of common arrangement of the LED lighting fixture and HVAC

air supply system for the energy-saving design. Figure 34 indicates how to utilize the

portion of LED heat harvested in the room space as a supplemental reheating power to re-

heat the discharged air in the conditioned room space.

Figure 34 Common arrangement of the LED lighting fixture and HVAC air supply system

Figure 35 LED heat harvested in the room space as a supplemental reheating power to re-heat

the discharged air in the conditioned room space

3.4.3 Energy Plus simulation results

In Energy Plus simulation, the input was the heat distribution pattern (‘conditioned

space/ceiling plenum split’) of the tested LED luminaires obtained from the laboratory

experiments in the Cold Room. The output was the building annual heating and cooling

energy consumption for the primary school classroom equipped with each of the four

different types of LED luminaires (ceiling-recessed or surfaced-mounted prototype LED

fixtures, two off-the-shelf commercial LED fixtures), respectively. For comparison, results

were obtained within/without deployment of the new energy-saving design. Tables 12

shows heating and cooling energy consumption of the typical primary school classroom

with four different LED luminaires (two prototype LED and two off-shelf LED). Figure 36

illustrate heating, cooling and total energy of the classroom with four LEDs in graphic.

Table 12 Annual heating and cooling energy for the primary school classroom

without deployment of the new lighting arrangement

Luminaire Ceiling Surface Ceiling recessed Surface recessed mounted commercial mounted prototype prototype LED LED commercial LED LED Heating power per 50 50 50 50 luminaire(W)* Heat distribution 60.4/39.6 69.6/30.4 47.2/52.8 58.0/42.0 (room/plenum split) Annual heating energy 55.88 55.31 56.65 56.00 (GJ)** Annual cooling energy 17.24 17.29 17.15 17.22 (GJ) Annual heating & cooling 73.12 72.60 73.80 73.22 energy (GJ) *Heating power of four luminaires are assumed to be 50W in energy plus simulation. **the annual heating energy contains energy consumption for space heating (in winter) and for reheating (in summer).

50 Cooling Energy (GJ) 40 Heating 30 Energy (GJ)

Total (GJ)

ENERGY CONSUMPTION (GJ) CONSUMPTION ENERGY 20

0 CEILING RECESSED SURFACE MOUNTED CEILING RECESSED SURFACE MOUNTED PROTOTYPE LED PROTOTYPE LED COMMERCIAL LED COMMERCIAL LED TYPE OF LED FIXTURES

Figure 36 Summary of Energy Plus simulation

Table 13 shows the annual heating and cooling energy consumption of the primary

school classroom with deployment of the new energy-saving design, which is only

compatible with the prototype LED in this study. The surface mounted prototype LED

luminaire trap 69.6% of heat in the room space, which is 9.2% more than the same LED

with ceiling recessed mounting method (60.4% in room space). Therefore, this 9.2% more

heat in the room space can be used for reheating in the summer and for heating in winter

by using surface mounted but result a little bit higher cooling energy consumption in

summer. As shown in Table 13 below, the new lighting arrangement is designed to utilize

harvested LED heat in the room space, and therefore the reheating power reduced by 50%

in theory (before deduction and after deduction in Table 13).

Table 13 heating and cooling energy of the classroom installed with the prototype LED

fixture and with new lighting arrangement deployed

Luminaires Ceiling recessed prototype Surface mounted prototype LED LED Heating power per luminaire(W) 50 50 Number of Luminaires in the 93.5푚2 Classroom 15 15 Heat distribution 60.4/39.6 69.6/30.4 Heating power in the room space (W/푚2) 4.92 6.00 Annual reheating energy (before deduction, in GJ) 4.93 4.70 Supplemental reheating energy (GJ) 2.47 2.68 Annual reheating energy (after deduction, in GJ) 2.46 2.02 Annual space heating energy (GJ) 50.95 50.61 Annual heating energy (GJ) 53.41 52.63 Annual cooling energy (GJ) 17.12 17.13 Annual heating and cooling energy (GJ) 70.53 69.76

CHAPTER 4

DATA ANALYSES AND RESULTS

Case 1 to Case 3 are designed to determine the relation between heating and cooling

energy consumption and the heat distribution pattern (ceiling/room split) of LED

luminaires. Case 4 shows how much energy can be saved by reducing reheat energy

consumption, which enabled by the prototype LED with integrated lighting arrangement.

4.1 Case 1: the ceiling surface-mounted prototype LED luminaire vs. the ceiling-

recessed prototype LED luminaire

Case 1 was the comparison of annual heating and cooling energy consumptions of

the primary school classroom equipped with ceiling surface-mounted prototype LED

luminaires vs. the ceiling recessed prototype LED luminaires. The results are shown in

Table 14. Based on the laboratory test results, the ceiling recessed prototype LED

luminaire could harvest 69.6% of its LED heat in the room space, which is 9.2% more than

that when the same LED luminaire was surface mounted. In Energy Plus simulation, the

primary school classroom equipped with surface-mounted prototype LED luminaires

consumed 72.60 GJ of building heating and cooling energy for one-year period, which is

0.52GJ less than that with the surface mounted prototype LED luminaires.

Table 14 Energy consumptions simulated in Energy Plus of the primary school

classroom equipped with surface mounted prototype vs. ceiling recessed prototype

Luminaire Ceiling recessed Surface mounted prototype LED prototype LED Heating power of luminaire 50 50 (W) Heat distribution (room/ceiling 60.4/39.6 69.6/30.4 split) Annual cooling energy 17.24 17.29 consumption of the primary school classroom (GJ) Annual heating energy 55.88 55.31 consumption of the primary school classroom (GJ) Annual heating and cooling 73.12 72.60 energy consumption (GJ)

4.2 Case 2: the ceiling recessed prototype LED luminaire vs. a ceiling recessed

commercial LED luminaire

Case 2 was the comparison of annual heating and cooling energy consumptions of the

primary school classroom equipped with the ceiling recessed prototype LED luminaire vs. a

ceiling recessed commercial LED luminaire. The results are shown in Table 14. It was found

that the ceiling recessed prototype LED luminaire with the new architecture of “Heat

Arrangement of LED Arrays in Low Profile” could harvest more LED heat in the room space

than the ceiling recessed commercial LED luminaire. Consequently, according to the Energy

Plus simulation results with the heat distribution pattern (‘conditioned space/ceiling plenum

split’) being the only variable, the primary school classroom installed with the ceiling recessed

prototype LED luminaire (which harvest 60.4% LED heat in room space) consumes less

heating and cooling energy per year than that with the ceiling recessed commercial LED

luminaire (harvested 47.2% LED heat in room space).

Table 15 Ceiling recessed prototype vs. ceiling recessed commercial LED luminaire

Luminaire Ceiling recessed Ceiling recessed

prototype LED commercial LED

Heating power of luminaire (W) 50 50

Heat distribution (room/ceiling split) 60.4/39.6 47.2/52.8

Annual cooling energy consumption of the primary 17.24 17.15

school classroom (GJ)

Annual heating energy consumption of the primary 55.88 56.65

school classroom (GJ)

Annual heating and cooling energy consumption (GJ) 73.12 73.80

4.3 Case 3: the surface mounted prototype LED luminaire vs. a surface mounted

commercial LED luminaire

Case 3 was the comparison of annual heating and cooling energy consumptions of the

primary school classroom equipped with the surface mounted prototype LED luminaire vs. a

surface mounted commercial LED luminaire. The results are shown in Table 15. The surface

mounted prototype LED luminaire with the new architecture design could harvest more LED

heat in the room space than the surface mounted commercial LED luminaire. Consequently,

according to the Energy Plus simulation results with heat the distribution pattern of

‘conditioned space/ceiling plenum split’ being the only variable, the primary school classroom

installed with the surface mounted prototype LED luminaire (harvested 69.6% of LED heat in

room space) consume less heating and cooling energy per year than that with the surface

mounted commercial LED luminaire (harvested 58% of heat in room space).

Table 16 Surface mounted prototype vs. surface mounted commercial LED luminaire

Luminaire Surface mounted Surface mounted

prototype LED commercial LED

Heating power of luminaire (W) 50 50

Heat distribution (room/ceiling split) 69.6/30.4 58.0/42.0

Annual cooling energy consumption of the 17.29 17.22

primary school classroom (GJ)

Annual heating energy consumption of the 55.31 56.00

primary school classroom (GJ)

Annual heating and cooling energy consumption 72.60 73.22

(GJ)

4.4 Case 4: the ceiling recessed and surface mounted prototype LED with energy

saving design and assumptions

Case 4 was designed to show how much energy can be saved by installing the new

prototype LED luminaires with the new integrative lighting and heating arrangement. The

results are shown in Tables 17 and 18. Table 16 is the comparison of energy consumptions

between the surface-mounted new prototype LED luminaires with the energy saving design

and the surface-mounted off-the-shelf LED luminaires. As shown in Table 17, compared

to the annual heating and cooling energy consumption of the primary school classroom

equipped with the surface mounted commercial LED luminaire, 4.7% of energy can be

saved by replacing those luminaires with the new prototype LED luminaire.

Table 17 Energy consumption of the classroom equipped with surface mounted prototype

LED with energy saving design vs. that of surface mounted commercial LED

Luminaire Surface mounted Surface mounted prototype LED commercial LED Heating power of luminaire (W) 50 50 Energy saving technology Utilizing room-portion heat N.A. as reheat power source Heat distribution (room/ceiling split) 69.6/30.4 58.0/42.0 Annual cooling energy consumption 17.13 17.22 of the primary school classroom (GJ) Annual heating energy consumption 52.63 56.00 of the primary school classroom (GJ) Annual heating and cooling energy 69.76 73.22 consumption (GJ)

Table 18 shows the energy consumption of the ceiling surface-mounted prototype

LED with the new energy saving design versus that of the conventional ceiling recessed

LED. Assuming the heating power is 50W for both types of luminaires (under

assumption#4), 4.4% annual heating and cooling energy could be saved in the primary

school classroom equipped with the ceiling recessed prototype LED luminaires with “new

lighting arrangement”, when compared to that of the same classroom equipped with the

ceiling recessed commercial LED luminaires.

Table 18 Energy consumption of the classroom equipped with ceiling recessed prototype

LED with energy saving design vs. that of ceiling recessed commercial LED

Luminaire Ceiling recessed prototype LED Ceiling recessed commercial LED Heating power of luminaire (W) 50 50 Energy saving Technology Utilizing room-portion heat as a N.A. supplemental reheat power source Heat distribution (room/ceiling split) 60.4/39.6 47.2/52.8 Annual cooling energy consumption 17.12 17.15 of the primary school classroom (GJ) Annual heating energy consumption 53.41 56.65 of the primary school classroom (GJ) Annual heating and cooling energy 70.53 73.80 consumption (GJ)

CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

The goal of this study was to verify the new system architecture of “Heat Arrangement of

LED Arrays in Low Profile” of future LED luminaires for reduction of building heating and cooling energy consumptions through integrative lighting and heating arrangement that interacts with the existing building HVAC system. This new system architecture of LED luminaire design enables more heat flux generated from the LED luminaires flowing forward to the conditioned room space while minimizing the portion trapped in the ceiling plenum. The portion of LED heat harvested in the room space can be utilized as supplemental space heating power in heating season

(in most cases), and also as supplemental reheating power in cooling season.

The prototype LED luminaire with the new system architecture, which was either ceiling recessed or surface mounted, and two other off-the-shelf commercial LED luminaires (one ceiling recessed and one surface mounted) were tested in the calorimeter in the Cold Room to obtain their heat distribution pattern in terms of ‘conditioned space/ceiling plenum split’. Based on the laboratory experiments in the Cold Room, among the tested samples of the LED luminaires, the new architecture of LED luminaire could help harvest more heat generated by LED chips in the room space than the other commercial LED luminaires do.

By analyzing Cases 1-3 in chapter 4, the following two statements can be concluded:

• Excluding the heat contribution from the LED driver, the prototype LED

luminaire with surface mounted design can help harvest 9.2% more heat

generated by LED chips in the room space than the same prototype

luminaire with ceiling recessed design do (in Case 1).

• Compared with the heat distribution pattern of ‘conditioned space/ceiling

plenum split’ of the two off-the-shelf LED luminaires bought from market,

the new integrative architecture of LED luminaire can help harvest more

heat generated by LED chips in the room space (in Case 2 and Case 3).

The annual heating and cooling energy consumption of a typical primary school classroom equipped with those tested LED luminaires were simulated in Energy Plus with details specified such as located in climate zone 4 (Lawrence, KS), single floor classroom with flat room and electric heating and cooling system, etc. Therefore, the results obtained in Energy Plus simulation in this study are all subject to those preconditions. By analyzing the heat distribution of LED luminaires and the simulation results of annual heating and cooling energy consumptions of the modeled primary school classroom with different LED luminaires installed in case 1 through case

3, two conclusions can be drawn as follows:

• With the assumption #4 that each LED fixtures are assigned with the same

total heating power in Energy Plus, higher percentage of heat harvested in

the conditioned room space can reduce annual heating and cooling energy

consumption of the primary school classroom.

• Higher percentage of heat harvested in the conditioned room space results

in higher annual cooling energy consumption and lower annual heating

energy consumption. With the harvested heating power of LED luminaire

increased in the room space, the amount of energy saved in the space

heating is larger than that of energy wasted in the space cooling.

Case 4 shows that the amount of energy saved by implementation of the integrated lighting and heating arrangement with the new prototype LED luminaires in the primary school classroom, by comparing to the same configurations without the new integrated lighting arrangement. By comparing the simulation results in Energy Plus of the primary school classroom installed with prototype LED luminaires (ceiling recessed and surface mounted) under the conditions described in assumption #3 (the prototype LED fixture with new lighting arrangement enable the heat generated by LEDs and harvested in room space can be utilizing as reheating power in cooling season) to that of the same space installed with current commercial LED luminaires (ceiling recessed and surface mounted), following conclusions can be drawn:

• The primary school classroom equipped with the new type of LED

luminaires with integrative lighting and heating arrangement can reduce

energy consumption for space heating, space cooling and reheating.

• The implementation of the integrative lighting and heating arrangement in

the primary school classroom equipped with the ceiling-recessed prototype

LED luminaires can save 4.7% of annual heating and cooling energy,

compared to the energy consumptions of the same classroom installed with

two commercial ceiling-recessed LED luminaire without adoption of the

integrative lighting and heating arrangement. The savings lowered to 4.4%

when compared to surface-mounted off-the-shelf LED luminaires.

5.2 Discussion

Since the heat distribution pattern (‘conditioned space/ceiling plenum split’) of LED luminaires affects building heating and cooling energy consumption, the luminaire form factor and its installation method become important in the development of future solid-state lighting luminaires. LED luminaires of next generation, which deploy the system architecture of “Heat Arrangement of LED Arrays in Low Profile” and the integrative lighting and heating arrangement with the existing HVAC system, may have most energy savings for space heating in winter and assisting the reheat coil of the HVAC system for reheating the discharged cool air in summer. How much cost reduction by using the new LED technologies may be estimated by the amount of energy saved (in GJ) and local electricity price ($/kWh).

On the other hand, both space heating and reheating could also use natural gas, which could be much cheaper than electricity. As a result, further investigation on the life cycle cost effectiveness of the integrative LED lighting and heating technologies is necessary.

Moreover, the heat gain from LED driver could be harvested together with the LED luminaire, which was not examined in the present study. In a typical case, a 25 Watts output

LED driver may generate approximately 3 Watts heating power, which takes 12% of its power output of the LED driver. Therefore, installation of LED drivers (e.g. attached or remote mounted; on the back or inside of luminaire’s housing) and its location (above or below the ceiling) could affect the heat distribution of a LED luminaire. Different LED drivers with different sizes and thermal specifications (heating power per unit output power) will surely change the total heating power of a LED luminaire.

Additionally, DOE (U.S. DoE, 2008, Figure 10) claimed at least 20% of energy consumption of LED luminaires is converted to radiant energy (visible light and radiation).

Therefore, the radiation/convection split, which was not measured in the present study, could influence the calculation of internal heat gains and finally affect building annual heating and cooling energy consumption. The radiation/convection split of LED luminaires is thus recommended to be measured in future studies, especially those studies on thermal interactions between new types of LED luminaires with integrative lighting and heating arrangement and conditioned supply air of the HVAC system.

REFERENCES

ASHRAE-RP-1282, 2018. Abstract, Lighting Heat Gain Distribution in Building, Oklahoma

State University.

Ball, H., & Green, D., 1983. The Impact of Lighting fixtures on Heating and Cooling Loads

Bardsley N., Bland S., Hansen M., Pattison L., Pattison M., Stober K., Yamada M. 2015. Solid-

State Lighting R&D Plan. May 2015. U.S. Department of Energy, Washington, DC.

Cai, H., 2015. Unpublished Research Plan & Proposal.

Cerbe & Wilhelms, Technische Thermodynamik, 2011. München: Carl Hanser Verlag.

Chung & Loveday, 1998. Numerical Modeling of Lighting/HVAC Interaction, Part 1: Light and

Heat Transfer in a Room.

Diederich, Simon, 2017. Development and Commissioning of a Calorimeter for Testing the

Thermal Performance of Luminaires.

Donald J. W, 1996. The Thermal Interactions between Lighting and HVAC Systems in Ceiling

and Floor Plenums.

Engineer Reference-Input and Output Reference, 2017. Available at https://www.

bigladdersoftware.com

Engineering Toolbox, Dec 2018. Available at: https://www.engineeringtoolbox.com.

Fisher & Chantrasrisalai, 2007. ASHRAE 1282-RP Final Report.

Handbook of Air Conditioning System Design, pp 1-99 to 1-100, 2016. Carrier Air Conditioning

Company, New York, 2017.

Hu, H & Heise, Z, 2013. A Field Study of Occupant Thermal Comfort with Radiant Ceiling Panel

systems.

IES handbook 10th Edition, 2011. Lighting for Education, Table 24.2.

Lawrence Berkeley national Laboratory, 1998. Interactions between Lighting and Space

Conditioning Energy Use in U.S. Commercial Buildings.

Lawrence local weather report, 2018. Available at: https://www.weather.com

Madison Gas and Electric Company, 2015. Reheat Systems for Commercial Buildings

Mitlas, G. P., 1974. Canadian Society for Mechanical Engineering Vol. 2, No. 3, 1973-74 p.169-

1974.

Navigant Consulting Inc., Feb 2010. Energy Savings Potential of Solid-State Lighting in General

Illumination Applications 2010 to 2030, U.S. Department of Energy.

Navigant Consulting Inc., Jan 2012. Energy Savings Potential of Solid-State Lighting in General

Illumination Applications, U.S. Department of Energy.

Rock, B. & Wolfe, D., 1997. A Sensitivity Study of Floor and Ceiling Plenum Energy Model

Parameters, ASHRAE Transactions.

Schultz J. & Johnson, B. 2013, Integration: Lighting and HVAC systems, Metro CD

Engineering, Dublin, Ohio.0

U.S. Department of Energy Building Energy Data Book, 2011. Retrieved Oct. 14, 2014.

U.S. Department of Energy, 2008. Updated per Multi-Year Program Plan figure 4-5, pg 58

U.S Energy Information Administration, 2014. Electricity consumption for lighting in US.

Wolfe, D., 1996. Thermal Interactions between Lighting and HVAC Systems in Ceiling and Floor

Plenums

Zhou, Lochhead, Zhong, and Huynh, 2018. Low Energy LED Lighting Heat Distribution in

Buildings. ASHRAE Research Project Final Report, RP-1681.

Appendix

Calculation sheet of the heating power of the tested LED luminaires in the Calorimeter

26.72

26.06

26.22

25.53

25.84

25.07

25.53

24.88

27.50

26.64

26.50

26.69

26.11

26.13

25.02

24.85

24.87 24.70

sensors

temperature temperature

Read from Read

(°C)

of air exhaust ofexhaust air Temperature Temperature

1.92

1.26

1.98

1.29

1.77

1.00

1.62

0.87

2.48

1.63

1.68

1.82

1.70

1.55

1.75

0.97

0.75

0.80 0.60

Calculated

(°C)

difference difference Temprature Temprature

24.80

24.24

24.07

23.91

24.01

25.02

24.82

24.87

24.41

24.58

24.39

24.05

24.10

24.07

24.10

sensors

temperature temperature

Read from Read

( of air intake ofintake air

Temperature Temperature

9.17

29.79

19.58

32.14

20.91

29.40

16.52

30.00

16.11

33.81

22.16

23.91

25.88

24.88

22.78

27.21

15.06

11.68 12.22

(W)

Power Power

Heating Heating Calculated

0.0294

(kW)

0.02979

0.01958

0.02091

0.01168

Power Power

Heating Heating

0.032137

0.016521

0.029996

0.016112

0.033808

0.022163

0.023905

0.025878

0.024885

0.022776

0.027213

0.027214

0.015056

0.012225 0.009169 Calculated

ROOM

CEILING

ROOM

CEILING

ROOM

CEILING

ROOM

CEILING

ROOM

CEILING

ROOM

CEILING

ROOM

CEILING

ROOM

CEILING

ROOM

CEILING

ROOM

CEILING

Room

Ceiling or Ceiling

Calorimet Cavity in Cavity

0.0139

0.01517

0.01515

0.015106

0.015083

0.015868

0.015847

0.016223

0.016201

0.018025

0.017999

0.013242

0.013221

0.013908

0.014267

0.014261

0.015165

0.015142

0.014903

0.014897

Calculated

(kg/s)

Flow Flow Mass Mass

0.01264

0.012948

0.013564

0.013873

0.015414

0.011329

0.011869

0.012331

0.013102

Calculated

(m3/s) Flow Flow

Volume Volume

1.68

1.76

1.80

2.00

1.47

1.54

1.60

1.70

1.64 1.64

Fan Sensors Fan

Read from Read

(m/s) Air Speed Speed Air

101.5920

101.9310

101.9306

100.9147

100.9147 102.6102

102.6102

weather station weather

Obtained from Obtained

Pressure (kPa) Pressure Barometric Barometric

0.0119

0.0128

0.0102

0.0110

0.0104

0.0112

0.0125

0.0134

0.0128

0.0138

0.0112

0.0115

0.0129

0.0132

0.0125

0.0126

0.0133

0.0136

0.0115

0.0117

Calculated Humidity

Absolute

0.600

0.647

0.515

0.557

0.526

0.568

0.634

0.678

0.648

0.699

0.568

0.586

0.648

0.661

0.625

0.634

0.666

0.681

0.587 0.600

sensor

humidity humidity

Read from Read

Humidity (%)Humidity

Relative Relative

Method

SUR-MONT

CEL-RESD

Mounting Mounting

1 10

Test Test Case 74